Olympus Pen F 38mm f/1.8 Review

fermy said:

Something tells me that Eric's chart will look about the same with lenses reversed and then one would have to chuck this down to sample variation .

Click to expand...

Something else I am going to do, get a good astigmatism chart. Anyone know of one?

I have a couple design ideas, just need to draw it up. Thinking something for subjective results, yes lens has slight astigmatism, moderate, heavy, etc. Maybe place the chart in the mid field and corner (center serves little to no purpose since there should be no astigmatism there).

Also, a decentered lens could have one or more really bad corners and one or more really good corners. So it doesn't hurt to at least glance at all 4 to make sure they are performing similarly.

Also, I take a number from Imatest that averages the horizontal and vertical resolutions and reports that. I believe it normalizes this to the vertical resolution.

Finally, I focus bracket around a focus point in the center. This lens has super severe field curvature so it is possible I didn't reach peak value for sharpness on corners of the lens. It was entertaining on the graph paper, when I focused on the center I could clearly see the field curving back from the center from the focus highlighting. This field curvature will cause hazing/glowing/poor contrast edges on the chart. My focus bracketing should cover this, but if the curvature was bad enough it might not. I could try specifically focusing at the corner to see if it will change the corner results at all.

Eric
--
I never saw an ugly thing in my life: for let the form of an object
be what it may - light, shade, and perspective will always make it
beautiful. - John Constable (quote)

See my Blog at: http://www.erphotoreview.com/ (bi-weekly)
Flickr Photostream: http://www.flickr.com/photos/28177041@N03/ (updated daily)

fermy said:

Something tells me that Eric's chart will look about the same with lenses reversed and then one would have to chuck this down to sample variation .

Click to expand...

Indeed. It also highlights the danger of drawing any meaningful conclusion from a comparison between single samples of any mass-produced lenses - whether visually or (especially) with numbers. After all, numbers don't lie, do they? ;-)

--
John Bean [GMT]

--
http://www.flickr.com/photos/lensman82/

John Bean (UK) said:

Indeed. It also highlights the danger of drawing any meaningful conclusion from a comparison between single samples of any mass-produced lenses - whether visually or (especially) with numbers. After all, numbers don't lie, do they? ;-)

Click to expand...

Sure they don't lie, it's interpretation that is always an issue. Anyway, I don't see any good reason for "especially" in your post: just as any owner's opinion, Eric's numbers provide description of reality that is relevant to one particular sample, only more objective. If you calculated MTF on your copies of Zuiko and Hexanon, the numbers would show the advantage of Zuiko; I hope you are not going to dispute that.

An alternative to drawing conclusions from single sample is drawing none at all, which is safer, but counterproductive. The more test results one has, the more reliable the conclusion is.

fermy said:

Sure they don't lie, it's interpretation that is always an issue. Anyway, I don't see any good reason for "especially" in your post: just as any owner's opinion, numbers provide description of reality that is relevant to one particular sample, only more objective.

Click to expand...

I wrote "especially" because numbers give more (but undeserved) credibility than a simple expression of opinion, even though they have no more reliability unless arrived at using a statistically meaningful sample size.

Numbers can provoke adjectives like "horrendous" from their interpreters where no horror actually exists.

--
John Bean [GMT]

John Bean (UK) said:

I wrote "especially" because numbers give more (but undeserved) credibility than a simple expression of opinion, even though they have no more reliability unless arrived at using a statistically meaningful sample size.

Click to expand...

Statistically meaningful sample size is unattainable in lens tests. Numbers deservedly give more credibility because unlike opinions they quantify the differences and express them in the known metric.

John Bean (UK) said:

Numbers can provoke adjectives like "horrendous" from their interpreters where no horror actually exists.

Click to expand...

When numbers started to be a prerequisite to "horrendous" comments?

One of us needs to put an end to this so I suppose it will have to be me.

I'm done here, others can decide for themselves from what has already been said and shown, both here and elsewhere.

Feel free to have the last word.

--
John Bean [GMT]

John Bean (UK) said:

One of us needs to put an end to this so I suppose it will have to be me.

I'm done here, others can decide for themselves from what has already been said and shown, both here and elsewhere.

Feel free to have the last word.

Click to expand...

Not sure why you are taking it that way, I thought we had a normal discussion.

viking79 said:

fermy said:

Something tells me that Eric's chart will look about the same with lenses reversed and then one would have to chuck this down to sample variation .

Click to expand...

Something else I am going to do, get a good astigmatism chart. Anyone know of one?

Click to expand...

A simple standard USAF chart measures radial and tangential resolution. Maybe you can print out a couple and place at strategic places on your board making it possible to let IMA-test do its things and then check the USAF patches manually without having to take another series of images.

viking79 said:

I have a couple design ideas, just need to draw it up. Thinking something for subjective results, yes lens has slight astigmatism, moderate, heavy, etc. Maybe place the chart in the mid field and corner (center serves little to no purpose since there should be no astigmatism there).

Click to expand...

Just remember you are checking a system. The CV35/1.2 corners improved when going from the Nex-5 to the 5N.

Also see my reply here: http://forums.dpreview.com/forums/read.asp?forum=1042&message=39994522

regards,

Jonas

John Bean (UK) said:

viking79 said:

John Bean (UK) said:

fermy said:

From the charts I see that expected sharpness-contrast for this lens is a bit weaker than that of Hexanon 40/1.8, a lens that I have. Is that different from your experience, John?

Click to expand...

Yes. I no longer have the Hexanon, but I found it unacceptably soft and "glowing" until beyond f/2.8. The Zuiko I have is as good wide open as the Hexanon was a full stop down, and is much better than the wide open Hexanon.

Your remark about "shockingly bad" corners is what prompted my first response, and if that is a valid judgement based on the numbers then the numbers are misleading you.

Click to expand...

I still have my Hexanon, so this begs a comparison shot I will see if I can't do something tonight or tomorrow.

Click to expand...

Here's a 100% crop composite of two shots of a Koren test chart, centred, all sharpening turned off, no black point adjustment, WB set from background:





One of them is the Hexanon 40/1.8, the other the Zuiko 38/1.8, both wide open. I leave it to you to guess which is which

PS: The Hexanon was lovely at f/2.8 and beyond, but the Zuiko walks all over it at wider apertures. I have no comparison corner tests I'm afraid.

Click to expand...

Talking wide open at f/1.8:

Just so we are clear, my Hexanon was sharper at the corners than the Zuiko wide open (resolution of the Zuiko is okay, but it has that hazyness to it that reduces MTF 50, edge goes from white, to light gray, slowly to black, but the edge itself is fairly sharp).

My center crop charts look nearly identical to yours. The Center of the Zuiko is much sharper than the Hexanon.

The corner shot crop, the hexanon is better, but the Zuiko could be fixed in processing a little. I probably prefer the performance of the Zuiko as I would rather have a sharper center wide open for available light people shots, etc, and sharper corners on stopping down (for landscape or similar).

Eric
--
I never saw an ugly thing in my life: for let the form of an object
be what it may - light, shade, and perspective will always make it
beautiful. - John Constable (quote)

See my Blog at: http://www.erphotoreview.com/ (bi-weekly)
Flickr Photostream: http://www.flickr.com/photos/28177041@N03/ (updated daily)

viking79 said:

My center crop charts look nearly identical to yours. The Center of the Zuiko is much sharper than the Hexanon.

The corner shot crop, the hexanon is better, but the Zuiko could be fixed in processing a little. I probably prefer the performance of the Zuiko as I would rather have a sharper center wide open for available light people shots, etc, and sharper corners on stopping down (for landscape or similar).

Click to expand...

Thanks for the confirmation Eric. I didn't doubt what my eyes told me but a second opinion never hurts

--
John Bean [GMT]

Has anyone made a side by side comparison of the older model (yellow tinted coating) with the newer one (blue tint)? Which is better?

absh said:

Has anyone made a side by side comparison of the older model (yellow tinted coating) with the newer one (blue tint)? Which is better?

Click to expand...

Wow, I did not know there was two versions! I have two copies of the 40mm pen f 1.4. I should put them side by side to see if there is a difference in tinting, and if so, then I can try make comparisons.

Eric, the only Olympus lens I have mounted onto my (also red) NEX-3 is the 15mm f8 body-cap lens. Want to trade? he-he!

Steve